Blood coagulation is a process consisting of a complex interaction of various blood components, or factors, which eventually give rise to a fibrin clot. Generally, blood components participating in the coagulation “cascade” are proenzymes or zymogens, i.e. enzymatically inactive proteins that are converted into an active form by action of an activator. Regulation of blood coagulation is largely accomplished enzymatically by proteolytic inactivation of the pro-coagulation factors Va and VIIIa achieved by activated protein C (APC) (Esmon, J Biol Chem 1989; 264; 4743–4746).
Protein C is a serine protease that circulates in the plasma as a zymogen with a half-life of approximately 7 hours and plasma levels are typically in the range of 3–5 mg/l. It is produced in vivo in the liver as a single chain precursor polypeptide of 461 amino acids. This polypeptide undergoes multiple post-translational modifications including a) cleavage of a 42 amino acid signal sequence; b) cleavage of lysine and arginine residues (positions 156 and 157) to make a two-chain inactive zymogen (a 155 amino acid light chain attached via a disulfide bridge to a 262 amino acid heavy chain); c) vitamin K-dependent carboxylation of nine glutamic acid residues of the light chain resulting in nine gamma-carboxyglutamic acid residues in the N-terminal region of the light chain; and d) carbohydrate attachment at four sites (one in the light chain and three in the heavy chain). Finally, the two-chain zymogen may be activated by removal of a dodecapeptide (the activation peptide) at the N-terminus of the heavy chain (positions 158–169) producing the activated protein C (APC).
Protein C is activated by limited proteolysis by thrombin in complex with thrombomodulin on the lumenal surface of the endothelial cell. As explained above, activation liberates a small 12 amino acid peptide (designated the activation peptide) from the N-terminal of the heavy chain. The APC has a half-life of approximately 15 minutes in plasma.
In the presence of its cofactor, protein S, APC proteolytically inactivates factors Va and VIIIa, thereby reducing thrombin generation (Esmon, Thromb Haemost 1993; 70; 29–35). Protein S circulates reversibly bound to another plasma protein, C4b-binding protein. Only free protein S serves as a cofactor for APC. Since C4b-binding protein is an acute phase reactant, the plasma levels of this protein varies greatly in many diseases and thus influence the anticoagulant activity of the protein C system.
The gene encoding human protein C maps to chromosome 2q13–q14 (Patracchini et al., Hum Genet 1989; 81; 191–192) spans over 11 kb, and comprises a coding region (exons II to IX) and a 5′ untranslatable region encompassing exon I. The protein domains encoded by exons II to IX show considerable homology with other vitamin K-dependent coagulation proteins such as factor IX and X. Exon II codes for a signal peptide, while exon III codes for a propeptide and a 38 amino acid sequence containing 9 Glu residues. The propeptide contains a binding site for the carboxylase transforming the Glu residues into dicarboxylic acid (Gla) able to bind calcium ions, a step required for phospholipid binding and protein C anticoagulant activity (Cheung et al., Arch Biochem Biophys 1989; 274; 574–581). Exons IV, V and VI encodes a short connection sequence and two EGF-like domains, respectively. Exon VII encodes both a domain encompassing a 12 amino acid activation peptide released after activation of protein C by thrombin, and the dipeptide 156–157 which, when cleaved off, yields the mature two-chain form of the protein. Exons VIII and IX encodes the serine protease domain.
The complete amino acid sequence of the human protein C has been reported by Foster et al., PNAS. USA 1986; 82; 4673–4677 and includes a signal peptide, a propeptide, a light chain, a heavy chain and an activation peptide.
Protein C binds to the endothelial cell protein receptor (EPCR). Binding of APC to EPCR renders APC incapable of inactivating factor Va and VIIIa, whereas binding of protein C to EPCR apparently enhances the activation rate of protein C by the thrombin-thrombomodulin complex. The physiological importance of these interactions is presently unknown. Apparently the binding of protein C to EPCR is strictly dependent on the presence of the Gla domain in a phospholipid independent manner (Esmon et al., Haematologica 1999; 84; 363–368).
APC is inhibited in the plasma by the protein C inhibitor as well as by alpha-1-antitrypsin and alpha-2-macroglobulin.
The experimental three-dimensional structure of human APC has been determined to 2.8 Å resolution and reported by Mather et al., EMBO J. 1996; 15; 6822–6831. They report the X-ray structure of APC in a Gla-domainless form. The structure includes a covalently bound inhibitor (D-Phe-Pro-Arg chloromethylketone, PPACK).
Protein C is currently isolated from prothrombin concentrates produced by monoclonal antibody affinity chromatography. Furthermore, protein C is produced recombinantly by expression from mammalian cells or modified protein C.
APC is used for the treatment of genetic and acquired protein C deficiency and is suggested to be used as anticoagulant in patients with some forms of Lupus, following stroke or myocardial infarction, after venous thrombosis, disseminated intravascular coagulation (DIC), septic shock, emboli such as pulmonary emboli, transplantation, such as bone marrow trans-plantation, burns, pregnancy, major surgery/traum and adult respiratory stress syndrome (ARDS).
Recombinant APC is produced by Eli Lilly and Co and phase III trials for the treatment of sepsis (Bernard et al., N Engl J Med (2001), 344, pp. 699–709) has recently been completed. Patients suffering from severe sepsis were given doses of 24 μg/kg/h for a total duration of 96 hours as infusion.
However, relatively high doses and frequent administration is necessary to reach and sustain the desired therapeutic or prophylactic effects of APC due to its short half-life. As a consequence adequate dose regulation is difficult to obtain and the need of frequent intravenous administrations of high levels of APC is problematic and expensive.
A molecule with a longer circulation half-life would decrease the number of necessary administrations and potentially provide more optimal therapeutic APC levels with concomitant enhanced therapeutic effect.
The circulation half-life of APC may be increased, e.g. as a consequence of reduced renal clearance, of reduced proteolytic degradation or reduced inhibition. This may be achieved, e.g., by conjugation APC to a non-polypeptide moiety, e.g. PEG or carbohydrates, capable of conferring a reduced renal clearance to the protein and/or effectively blocking proteolytic enzymes or inhibitors from physical contact with the protein. Furthermore, this may also be achieved by mutating the protein C molecule in such a way that it remains active but blocks the binding of inhibitors to the protein.
PEGylated wild-type APC is described in JP 8-92294.
WO 91/09960 discloses a hybrid protein comprising modifications in the heavy chain part of protein C.
WO 01/59084 describes protein C variants comprising the substitutions D167F+D172K in combination with at least one further substitution in position 10, 11, 12, 32, 194, 195, 228, 149, 254, 302 or 316. The variants disclosed in WO 01/59084 are stated to have an increased anticoagulant activity.
WO 98/44000 broadly describes protein C variants with an increased amidolytic activity.
EP 0 323 149 describes zymogen forms of protein C with the following mutations in the heavy chain: D167F/G/Y/W. Such variants are stated to have an increased sensitivity to activation by thrombin.
WO 00/66754 reported that substitution of the residues naturally occurring in the positions 194, 195, 228, 249, 254, 302 or 316 lead to an increased half-life of APC in human blood as compared to the wild-type APC. The variants disclosed in WO 00/66754 are not within the scope of the present invention.
WO 99/63070 describes a C-terminally truncated form of protein C.
EP 0 946 715 reported chimeric protein C polypeptides where the protein C Gla domain was replaced by Gla domains from other vitamin K-dependent polypeptides, such as factor VII, factor X and prothrombin.
WO 99/20767 and WO 00/66753 discloses vitamin K-dependent polypeptide variants containing modifications in the Gla domain.
U.S. Pat. No. 5,453,373 discloses human protein C derivatives which have altered glycosylation patterns and altered activation regions, such as N313Q and N329Q. The variants disclosed in U.S. Pat. No. 5,453,373 are not within the scope of the present invention.
U.S. Pat. No. 5,460,953 discloses DNA sequences encoding zymogen forms of protein C, which have been engineered so that one or more of the naturally occurring glycosylation sites have been removed. More specifically, U.S. Pat. No. 5,460,953 discloses the variants N97Q, N248Q, N313Q and N329Q. The variants disclosed in U.S. Pat. No. 5,460,953 are not within the scope of the present invention. None of the disclosed variants in any of the above-mentioned prior art references are within the scope of the present invention.
U.S. Pat. No. 5,270,178 is directed to specific protein C variants, wherein I 171 is deleted and wherein Asp is replaced by Asn.
U.S. Pat. No. 5,041,376 relates to a method for identifying and shielding functional sites or epitopes of transportable proteins, wherein additional N-linked glycosylation site(s) have been introduced.
U.S. Pat. No. 5,766,921 is directed to protein C variants having increased resistance to inactivation by human plasma or α1-antitrypsin, where the heavy chain contains substitutions from the corresponding bovine heavy chain.
WO 01/57193 reports a protein C variant comprising a double mutation, one mutation in positions 10, 11, 32 or 33 and one mutation in positions 194, 195, 228, 249, 254, 392 or 316.
WO 01/36462 relates to protein C variants comprising a substitution in position 12, optionally combined with substitutions in positions 10 and/or 11.
WO 00/26354 is directed to a method for producing glycosylated protein variants having reduced allergenicity.
WO 00/26230 is directed to a method for selecting a protein variant having reduced immunogenecity.
The DNA sequence and the corresponding amino acid sequence of human wild-type protein C, including the precursor form thereof, is disclosed in inter alia U.S. Pat. No. 4,775,624 and U.S. Pat. No. 4,968,626.
None of the variants disclosed in any of the above-identified patents/patent applications are within the scope of the present invention.